كليدواژه :
Nanofluid , Thermal conductivity , Carbon , based , QSPR , CORAL.
چكيده فارسي :
Heat transfer processes consume much of the total energy consumption, so increasing their efficiency is essential. In this area, nanofluids have attracted increasing attention to achieve this goal. on the other hand, considering repeated experimental studies can be toxic, time-consuming, and costly, it is necessary to focus on theoretical developments. The project used the quantitative structure-property relationship (QSPR) method to predict carbon nanofluids thermal conductivity. were checked, on the factors of temperature, concentration, and specific surface area on enhancing the thermal conductivity of carbon nanofluids. Our dataset contains 72 data points. Data on graphene nanoplatelets (GNP) dispersed in distilled water were collected and modeled using the Monte Carlo method in a software called CORAL and defined the structure using quasi-SMILES. This method was used to calculate the optimal descriptors. The specifications of the proposed models were fully calculated, and the results indicate that the developed model is statistically significant for predicting the thermal conductivity ratio of carbon nanofluids. R2m all splits more than 0.5 and ΔR2m less than 0.2. The best model is related to split 2 with R2m=0.9290 and ∆R2m=0.0301The best model is related to Split 2 with R2m=0.9290 and ∆R2m=0.0301, while the weakest statistical performance was observed in Split 4 with R2m=0.6509 and ∆R2m=0.1890. The application domain of the created models was clearly defined and correlation weights were calculated for all splits. The results showed that high concentrations, high temperatures, and high surface areas have a positive effect on the thermal conductivity ratio, while low temperatures have a negative effect on the target properties. The statistical performance of the models was satisfactory.
